A Proposal for Creation of Beverage Suited for User by Blending Juices based on Interactive Genetic Algorithm

Obtaining media contents suited to user’s preference and kansei is one of the important targets of kansei engineering (affective engineering). However, it is still difficult, because preference and kansei are quite different between users and they are difficult to describe as equation or function. As a creation method of a beverage, this study proposed Interactive Genetic Algorithm that blends source juices and creates new beverages suited to each user. A target of optimization was the ratio of the source juices. A concrete system was constructed with personal computer, Arduino, bottles containing source juices, and water pumps. To show efficiencies of the proposed IGA, a tasting experiment composed of two steps was conducted. In the search step, the subjects evaluated blended beverages as GA individuals. As results, continuous increase in the fitness values was observed and it was significant. In the progress of the ratios, different and common trends were observed. However, significant increase in the fitness value was not observed in the evaluation step.